Update

[ghstack-poisoned]

.github/scripts/github_utils.py

@@ -3,14 +3,12 @@
 import json
 import os
 import warnings
-
 from dataclasses import dataclass
-from typing import Any, Callable, cast, Dict, List, Optional, Tuple, Union
+from typing import Any, Callable, Dict, List, Optional, Tuple, Union, cast
 from urllib.error import HTTPError
 from urllib.parse import quote
 from urllib.request import Request, urlopen
 
-
 GITHUB_API_URL = "https://api.github.com"
 
 

.github/scripts/gitutils.py

@@ -9,14 +9,14 @@
 from typing import (
     Any,
     Callable,
-    cast,
     Dict,
     Iterator,
     List,
     Optional,
     Tuple,
     TypeVar,
     Union,
+    cast,
 )
 
 T = TypeVar("T")
@@ -45,7 +45,7 @@ def fuzzy_list_to_dict(items: List[Tuple[str, str]]) -> Dict[str, List[str]]:
 
 
 def _check_output(items: List[str], encoding: str = "utf-8") -> str:
-    from subprocess import CalledProcessError, check_output, STDOUT
+    from subprocess import STDOUT, CalledProcessError, check_output
 
     try:
         return check_output(items, stderr=STDOUT).decode(encoding)

.github/scripts/label_utils.py

@@ -1,11 +1,10 @@
 """GitHub Label Utilities."""
 
 import json
-
 from functools import lru_cache
-from typing import Any, List, Tuple, TYPE_CHECKING, Union
+from typing import TYPE_CHECKING, Any, List, Tuple, Union
 
-from github_utils import gh_fetch_url_and_headers, GitHubComment
+from github_utils import GitHubComment, gh_fetch_url_and_headers
 
 # TODO: this is a temp workaround to avoid circular dependencies,
 #       and should be removed once GitHubPR is refactored out of trymerge script.

.github/scripts/trymerge.py

@@ -23,44 +23,41 @@
 from typing import (
     Any,
     Callable,
-    cast,
     Dict,
     Iterable,
     List,
     NamedTuple,
     Optional,
     Pattern,
     Tuple,
+    cast,
 )
 from warnings import warn
 
 import yaml
 from github_utils import (
+    GitHubComment,
     gh_fetch_json_list,
     gh_fetch_merge_base,
     gh_fetch_url,
     gh_graphql,
     gh_post_commit_comment,
     gh_post_pr_comment,
     gh_update_pr_state,
-    GitHubComment,
 )
-
 from gitutils import (
+    GitRepo,
     are_ghstack_branches_in_sync,
     get_git_remote_name,
     get_git_repo_dir,
-    GitRepo,
     patterns_to_regex,
     retries_decorator,
 )
 from label_utils import (
     gh_add_labels,
     gh_remove_label,
-    has_required_labels,
-    LABEL_ERR_MSG,
 )
-from trymerge_explainer import get_revert_message, TryMergeExplainer
+from trymerge_explainer import TryMergeExplainer, get_revert_message
 
 # labels
 MERGE_IN_PROGRESS_LABEL = "merging"
@@ -1477,7 +1474,7 @@ def checks_to_str(checks: List[Tuple[str, Optional[str]]]) -> str:
 
 
 def checks_to_markdown_bullets(
-    checks: List[Tuple[str, Optional[str], Optional[int]]]
+    checks: List[Tuple[str, Optional[str], Optional[int]]],
 ) -> List[str]:
     return [
         f"- [{c[0]}]({c[1]})" if c[1] is not None else f"- {c[0]}" for c in checks[:5]
@@ -1716,7 +1713,7 @@ def get_readable_drci_results(drci_classifications: Any) -> str:
         try:
             print(f"From Dr.CI checkrun summary: {drci_summary}")
             drci_classifications = json.loads(str(drci_summary))
-        except json.JSONDecodeError as error:
+        except json.JSONDecodeError:
             warn("Invalid Dr.CI checkrun summary")
             drci_classifications = {}
 
@@ -1887,7 +1884,6 @@ def do_revert_prs(
     dry_run: bool = False,
 ) -> None:
     # Prepare and push revert commits
-    commit_shas: List[str] = []
     for commit_sha, pr in shas_and_prs:
         revert_msg = f"\nReverted {pr.get_pr_url()} on behalf of {prefix_with_github_url(author_login)}"
         revert_msg += extra_msg

.github/scripts/trymerge_explainer.py

@@ -2,7 +2,6 @@
 import re
 from typing import List, Optional, Pattern, Tuple
 
-
 BOT_COMMANDS_WIKI = "https://github.com/pytorch/pytorch/wiki/Bot-commands"
 
 CIFLOW_LABEL = re.compile(r"^ciflow/.+")

.github/workflows/float8_test.yml

@@ -35,6 +35,7 @@ jobs:
       runner: ${{ matrix.runs-on }}
       gpu-arch-type: ${{ matrix.gpu-arch-type }}
       gpu-arch-version: ${{ matrix.gpu-arch-version }}
+      submodules: recursive
       script: |
         conda create -n venv python=3.9 -y
         conda activate venv

.github/workflows/nightly_smoke_test.yml

@@ -31,6 +31,7 @@ jobs:
       runner: ${{ matrix.runs-on }}
       gpu-arch-type: ${{ matrix.gpu-arch-type }}
       gpu-arch-version: ${{ matrix.gpu-arch-version }}
+      submodules: recursive
       script: |
         python -m pip install --upgrade pip
         pip install ${{ matrix.torch-spec }}

.github/workflows/regression_test.yml

@@ -40,6 +40,7 @@ jobs:
       runner: ${{ matrix.runs-on }}
       gpu-arch-type: ${{ matrix.gpu-arch-type }}
       gpu-arch-version: ${{ matrix.gpu-arch-version }}
+      submodules: recursive
       script: |
         conda create -n venv python=3.9 -y
         conda activate venv
@@ -93,6 +94,7 @@ jobs:
       runner: ${{ matrix.runs-on }}
       gpu-arch-type: ${{ matrix.gpu-arch-type }}
       gpu-arch-version: ${{ matrix.gpu-arch-version }}
+      submodules: recursive
       script: |
         conda create -n venv python=3.9 -y
         conda activate venv

.gitmodules

@@ -0,0 +1,3 @@
+[submodule "third_party/cutlass"]
+	path = third_party/cutlass
+	url = https://github.com/NVIDIA/cutlass

.pre-commit-config.yaml

@@ -26,4 +26,5 @@ repos:
         alias: ruff-isolated
         args:
           - --isolated
-          - select F821,F823,W191
+          - --select
+          - F821,F823,W191

benchmarks/bench_galore_fused_kernels.py

@@ -8,9 +8,7 @@
 def run(args):
     dtype = getattr(torch, args.dtype)
     allow_tf32 = args.allow_tf32
-    fp8_fast_accum = False
     torch.backends.cuda.matmul.allow_tf32 = allow_tf32
-    kernel = args.kernel
     M, N = args.M, args.N
     rank = args.rank